[<= Back]


Use of the SWISS-MODEL server: Homology Modeling of a 3D Structure from an Aminoacid Sequence

Example.

Step-by-step example of the use of the "SWISS-MODEL First Approach mode" to obtain a 3D model of a sequence.


Sequence input.

Starting from a lineal aminoacid sequence:
>unknown
MSKVPRNFRL LEELEKGEKG FGPESCSYGL ADSDDITMTK WNGTILGPPH SNHENRIYSL
SIDCGPNYPD SPPKVTFISK INLPCVNPTT GEVQTDFHTL RDWKRAYTME TLLLDLRKEM
ATPANKKLRQ PKEGETF


The first step is to introduce the sequence into the "SWISS-MODEL First Approach mode" web server and to launch the process [See Figure].


Recepction of a confirmatory mail.

SwissModel send us a confirmatory mail indicating the date, title of the search, identification code and somo other information... Access to the example mail.


Recepction of a mail containing different processes followed by SWISS-MODEL.

This second message shows a description of the processes followed by the  SWISS-MODEL program (e.g. Access to the processes mail).:
 
  1. SWISS-MODEL perform a homology searching on a 3D structure database using BLASTP2 and the database "ExNRL-3D". The results are sorted according to the BLAST P(N) score:
  2.      AlignMaster output
         ============================================================
         
         Length of target sequence: 137 residues
         
         Searching sequences of known 3D structures
         Found 12UCE.pdb with P(N)=5.0e-12
         Found 11AYZ.pdb with P(N)=1.2e-08
         Found 11AAK.pdb with P(N)=5.3e-08
         Found 12AAK.pdb with P(N)=5.3e-08
         Found 11UCZ.pdb with P(N)=1.1e-07
         Found 12UCZ.pdb with P(N)=1.1e-07
         Found 11A3S.pdb with P(N)=1.0e-05
  3. The  "SIM" program selects all the templates with a sequence identity higher that 25% and an aligned region over 20 residues:
  4.      Extracting template sequences
         
         Running pair-wise alignments with target sequence
         Sequence identity of templates with target:
         
         12UCE.pdb: 22.85 % identity
         11AYZ.pdb: 27.8 % identity
         11AAK.pdb: 17.15 % identity
         12AAK.pdb: 17.15 % identity
         11UCZ.pdb: 29.6 % identity
         12UCZ.pdb: 29.6 % identity
         11A3S.pdb: 22.5 % identity
         
         Looking for template groups
         Global alignment overview:
         
         Taget Sequence: |====================================================================|
         12UCE.pdb       |                  -------------------------------------------------
         11AYZ.pdb       |                    ------------------------------------------------
         11AAK.pdb       |               -----------------------------------------------------
         12AAK.pdb       |               -----------------------------------------------------
         11UCZ.pdb       |                 --------------------------                        
         12UCZ.pdb       |                 --------------------------                        
         11A3S.pdb       |                 --------------------------------------------------
         
         AlignMaster found 1 regions to model separately:
                 1: Using template(s)   11A3S.pdb 11AAK.pdb 11AYZ.pdb 11UCZ.pdb 12AAK.pdb 12UCE.pdb 12UCZ.pdb
         
         12UCE.pdb has been rejected,  too low similarity with Target sequence (22.85 % identity.)
         11AAK.pdb has been rejected,  too low similarity with Target sequence (17.15 % identity.)
         12AAK.pdb has been rejected,  too low similarity with Target sequence (17.15 % identity.)
         11A3S.pdb has been rejected,  too low similarity with Target sequence (22.5 % identity.)
  5. In this step the "ProModII" program input file is generated. This program, using the templates from the "ExPDB" database, generates all the models from the input sequence:
  6.      Creating Batch files for ProMod (if any):
                 Batch.1: residues 30 - 137 of submitted sequence.
         
         Exiting AlignMaster
         
         
         ProModII trace log for Batch.1
         ============================================================
         
         ProModII: Loading Template: 11AYZ.pdb
         ProModII: Loading Template: 11UCZ.pdb
         ProModII: Loading Template: 12UCZ.pdb
         ProModII: Loading Raw Sequence
         ProModII: Iterative Template Fitting
         ProModII: Iterative Template Fitting
         ProModII: Generating Structural Alignment
         ProModII: Aligning Raw Sequence
         ProModII: Refining Raw Sequence Alignment
         ProModII: Weighting Backbones
         ProModII: Averaging Sidechains
         ProModII: Adding Missing Sidechains
         ProModII: Small Ligation (C-N < 3.0A) ignored;
         ProModII: GROMOS will repair it at residue ASP 73
         ProModII: Building CSP loop with anchor residues THR 60 and GLU 63
         ProModII:   Number of Ligations found:  (1)
         ProModII: all loops are bad; continuing CSP with larger segment
         ProModII: Building CSP loop with anchor residues PRO 59 and GLU 63
         ProModII:   Number of Ligations found:  (10)
         ProModII: ACCEPTING loop    5: clash=   1 FF=         53.8 PP=-18.41
         ProModII: Dumping Preliminary Model
         ProModII: Dumping Sequence Alignment
         ProModII: Done.
  7. In the last step, the "Gromos96" program optimizes the model(s) using a energy minimization process:
  8.      Gromos96 trace log for Batch.1
         ============================================================
         
         Now running PROCS1 on file batch-procs0.dat ... Done.
         Now running PROCS2 on file batch-procs1.dat ... Done.
         Now running PROGMT on file batch-procs2.dat ... Done.
         Now running PROGCH on file batch-procs2.dat ... Done.
         Now running PROMD on file batch-progch.dat ... Done.
         Now running PROMD on file batch-promd0.dat ... Done.
         Detection of SS-Bonds within batch ...

Recepction of a mail with the 3D coordinates of the model.

Finally, the SWISS-MODEL server send a mail containing the 3D coordinates of the model based on the homology of the initial sequence with the ones of known spatial structure. (e.g. Access to the coordinates mail.). Is important to realize that this coordinates are from a predictive process and, thus, they are not 100% accurate.
 


the model



Juan Antonio García Ranea. Protein Design Group. Centro Nacional de Biotecnología (C.N.B. - C.S.I.C.) Campus Universidad Autonoma. Cantoblanco. 28049 Madrid. Tlf: +34-91-5854669. Fax: +34-91-5854506. e-mail: garcia@gredos.cnb.uam.es

PGP-17-10-00